News for hackers. Stuff that matters.
Re:Linus Torvalds: 'I'll never be cuddly but I can be more p... by DFHippie on Thursday September 27, 2018 @10:15AM (#18084650)
It is important to note that this isn't merely a matter of opinion. One can study the effect/presence of unconscious bias objectively and quantitatively. These studies have large error bars but they use conventional and accepted statistics.

To the extent it is politicized, it's politicized the way evolution and climate change are: groups who are politically opposed have opinions about the consequences of these studies. One or the other group sometimes finds empirical reality to be in conflict with the arguments they wish to make in support of their preferred policies.

Re:Linus Torvalds: 'I'll never be cuddly but I can be more p... by mooseburger on Thursday September 27, 2018 @02:15PM (#18087216)
As far as I can tell, unconscious bias has fallen victim to the replication crisis.

https://www.thecut.com/2017/01/psychologys-racism-measuring-...

Re:Linus Torvalds: 'I'll never be cuddly but I can be more p... by pseudalopex on Thursday September 27, 2018 @02:43PM (#18087455)
The article seems to focus on the Implicit Association Test. Other studies have used devices like identical job applications with different names.
Re:Linus Torvalds: 'I'll never be cuddly but I can be more p... by gdy on Thursday September 27, 2018 @05:32PM (#18088940)
Is there a meta-study of Cochrane quality on this subject? Because, you know, replication crisis.
Re:Linus Torvalds: 'I'll never be cuddly but I can be more p... by beaconstudios on Thursday September 27, 2018 @10:22AM (#18084742)
note that I didn't suggest that the classical liberal side deny the existence of subconscious bias. Just the scale of the impact of said bias, and what can realistically be done about it (or, perhaps, the cost of attempting to correct it). There are probably people who deny the existence of bias and you're right in that they don't express a reasonable opinion.
Re:Linus Torvalds: 'I'll never be cuddly but I can be more p... by tenaciousDaniel on Thursday September 27, 2018 @10:26AM (#18084795)
Taken at face value, your statement seems to imply that if there is any research devoted to a topic using "conventional and accepted statistics", then no one can have an opinion on it.
Re:Linus Torvalds: 'I'll never be cuddly but I can be more p... by DFHippie on Thursday September 27, 2018 @10:45AM (#18084987)
I don't think you're taking it at face value. Note the word "merely".

There are typically many arguments for and against a particular policy, many costs and benefits. These costs and benefits don't hit everyone equally, which explains most of the difference in political opinion. One category of argument in the political debate is to claim that particular costs or benefits are illusory or of a different scale. The empirical evidence can show that this is false. This does not eliminate the debate. One would hope it would then proceed on different arguments, because denying empirical reality makes any resolution other than physical domination of one party by the other impossible. But even if it does, costs and benefits still fall unequally on the participants, meaning the political division remains. This difference in opinion is normal and legitimate.

[ETA I do not think it is pure coincidence that the group that believes it is more likely to win a contest of physical dominance is the one more likely to deny the validity of empirical evidence.]